You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This is very similar to the problem reported in #2266, except it is surfaced in a way that the fix for that issue doesn't work.
The way it surfaced is I modified the spec with a toleration that is not allowed. This propagated to the statefulset which then resulted in the error pasted below. I then fixed the spec by removing this toleration, but eck is not propagating the change to the statefulset anymore.
Events:
Type Reason Age From Message
---- ------ ---- ---- -------
Warning FailedCreate 6m18s (x32 over 3h59m) statefulset-controller create Pod es-test-2 in StatefulSet es-test failed error: admission webhook "{my-auth-service}" denied the request: {my-error-message — related to a toleration i set in the spec that is not allowed}
The problem here is that the "dry-run" solution in #3020 (which works in most cases) fails here because the failure does not occur in the pod but rather in the statefulset controller.
Is this an issue the community is interested in? I think a fix here could be to implement the timeout described in #2266 (comment). This would still be nice to have on top of the existing "dry-run" validation.
The text was updated successfully, but these errors were encountered:
Bug Report
This is very similar to the problem reported in #2266, except it is surfaced in a way that the fix for that issue doesn't work.
The way it surfaced is I modified the spec with a toleration that is not allowed. This propagated to the statefulset which then resulted in the error pasted below. I then fixed the spec by removing this toleration, but eck is not propagating the change to the statefulset anymore.
The problem here is that the "dry-run" solution in #3020 (which works in most cases) fails here because the failure does not occur in the pod but rather in the statefulset controller.
Is this an issue the community is interested in? I think a fix here could be to implement the timeout described in #2266 (comment). This would still be nice to have on top of the existing "dry-run" validation.
The text was updated successfully, but these errors were encountered: